Transistor amplifiers have a peak efficiency for a particular input power that is a function of geometry (i.e. circuit components and layout), load and supply voltage. In conventional radio frequency (RF) power amplification these characteristics are fixed based on the peak input level expected. For amplifiers presented with an input signal having a wide dynamic range, the input signal infrequently achieves peak levels and frequently operates below peak levels. As such, the amplifier may exhibit low overall efficiency.
A solution to the problem of low amplifier efficiency is to vary one or more of the above-stated characteristics (geometry, load, supply voltage) in response to the input signal. Techniques to vary one or more of these characteristics are known in the art.
Techniques that vary the device geometry and load tend to be very dependent on the particular power amplifier topology used, and generally present challenging RF problems. Repeatability of such designs in production is generally a problem.
Various techniques are known in the art for enhancing amplifier efficiency based on the supply voltage. Of supply voltage based efficiency enhancement schemes, there are two broad classifications of solution. These solutions are:
(i) envelope elimination and restoration, and
(ii) envelope tracking.
Envelope elimination and restoration requires the amplifier to be driven saturated, and all the envelope information to be applied through the amplifier supply. This technique tends to be generally too demanding upon the supply modulator when using high modulation bandwidths, and thus has limited usefulness in practical applications.
With envelope tracking, the amplifier is driven in a substantially linear fashion. Envelope tracking requires an efficient power supply capable of delivering high modulation power bandwidths. In known techniques, a switched mode pulse width modulator (commonly referred to as class S) is used to realise an efficient variable supply to the power amplifier. However, in order to operate at full bandwidth, the supply must switch at many times the bandwidth of the modulation, and this excessively high switching speed results in poor modulator efficiency.
In another prior art envelope tracking technique, a plurality of highly efficient intermediate power supplies are provided, and the power supplies are switched as required by the envelope level. This switching creates transient disturbances that degrade the spectrum with high order intermodulation products, and makes linearisation difficult by introducing supply dependent non-linearities alongside input dependent non-linearities.
In a further modification to this technique, the switching of the power supplies is combined with a linear amplifier to provide a smooth transition between switch levels and remove the supply dependent linearisation requirement. The aim of this form of envelope tracking is to provide a unique value of supply voltage for every envelope level. However, there is a problem in achieving this without impact upon tracking speed capability.
A variable level power supply must be able to switch between different supply levels in order to provide the necessary varied voltage supply levels. One known method of achieving this is to provide a means of coarse switching between a number of voltage sources, However this course switching results in errors in the voltage supply signal, as illustrated in FIG. 1. Referring to FIG. 1, reference numeral 206 denotes a dashed line representing the idealised envelope of the power supply voltage. This idealised power supply envelope 106 preferably tracks the envelope of the input signal to a device which the power supply is driving, such as a power amplifier. The reference numeral 102 denotes a line representing the envelope of the input signal to the device.
However, in practice, using coarse switching, the power supply envelope follows a shape as represented by the stepped curve 104. In the example illustrated with reference to FIG. 1, it is assumed that the switched supply has four coarse levels V1 to V4. As the envelope of the input signal to the input signal to the device, 102, reaches the one of the voltage levels V1 to V4, the supply voltage is appropriately switched. As can therefore be seen in FIG. 1, the supply voltage switches between four levels. As such, there are portions of the cycle where the supply voltage is excessive, and is therefore in error. As illustrated by the hatched area 108, the stepped supply voltage implementation gives rise to inefficiencies, as the hatched area 108 represents wasted energy.
In order to address this problem, it is known to sum the selected coarse voltage with a finely adjustable voltage source in order to provide interpolation, and to minimise the error.
A particularly advantageous technique for controlling the selection of the supply voltage, and adjustments thereto, to improve efficiency is taught in British patent application publication number 2398648.
It is an aim of the present invention to provide an improved scheme for summing voltages in the generation of a supply voltage.